A Genome Wide-Association Study of Recurrent Laryngeal Neuropathy in Belgian Draft Horses

Recurrent Laryngeal Neuropathy (RLN) is a major cause of poor performance in athletic horses. Prevalence in Belgian Draft Horses (BDH) can be as high as 41%. Our aim was to investigate the genetic risk factors associated with RLN in BDHs through a genome wide association study (GWAS).

RLN-affected (grade >3/4: n=39) and control (grade 1/4: n=19) BDHs were genotyped on the Illumina Equine SNP50 BeadChip. Genotypes were imputed to a uniform set of 2 million SNPs with BEAGLE using genotype data from 192 additional BDHs. Following quality control and data pruning in PLINK, 51581 SNPs (GWAS) and 1157722 SNPs (imputed) remained for analysis. Haplotype analysis was performed across any regions of interest using haplo.stats in R. Whole genome sequence data (4 cases/4 controls) was used to investigate these regions of interest, and alleles prioritized using Phenolyzer and Endeavour, using a seed gene list from similar human peripheral neuropathies.

No SNPs reached genome wide significance, based upon Bonferroni correction (GWAS: p=2.05x10^-6, imputation: p=4x10^-8). Several SNPs on chr3 were most associated with RLN; GWAS- ECA3.91784157 (p=7.96x10^-6), ECA3.94418288 (p=4.57x10^-5) and Imputed- ECA3.91486401-91493210 (p=2.22x10^-5). Haplotype association from ECA3.91352673-110131522 revealed significant association across a span of 5 SNPs (p values; 0.0036-0.0041).

A total of 73 genes within chr3 87-110Mbp were prioritized, and 124086 variants were seen across the chr3 region. Allele frequency differences, in this region, between cases and controls, in WGS data, have not yet revealed any obvious functional perturbations in the top 10 prioritized genes. Further analysis is continuing.